The amyloid hypothesis of Alzheimer disease (AD) pathogenesis has been the leading theory behind our understanding of the disease mechanism and the predominant therapeutic target for the past decade or more. 1 The rationale for believing that b-amyloid (Ab) is critical to the etiology of AD is simple and compelling. Very high levels of an amyloidogenic peptide, Ab, accumulate in the brains of patients with AD; all genetic forms of AD are associated with increased amyloid deposition, including the early onset autosomaldominant disease associated with mutations in presenilin 1 and 2 or the amyloid precursor protein (APP) itself. In addition, AD develops in nearly all individuals with Down syndrome, who have a triplication of the APP gene. The APOE4 polymorphism, which markedly increases the risk for late onset AD, also is associated with an increase in amyloid deposits, whereas the APOE2 polymorphism that diminishes risk for AD is associated with decreased amyloid levels. Conversely, a rare protective polymorphism in the APP gene itself is associated with decreased Ab synthesis. 2–7 Thus, a primary, causative, pathogenic link between Ab and AD seems secure.

Nonetheless, there are some twists that complicate the picture. Ab accumulation does not correlate well with extent of neuronal loss or cognitive dysfunction, 8, 9 and demonstrating direct neurotoxicity of Ab has been difficult in most animal models, suggesting the possible existence of key intermediates between amyloidosis and neurodegeneration. The most obvious candidates as intermediates are neurofibrillary tangles that contain tau; such tangles accumulate in AD brains, but in contrast to amyloid, their levels correlate quantitatively with the severity of clinical symptoms and with the areas of the brain thought to be responsible for the specific symptoms of the disease, such as memory impairment. Despite this, Ab has remained the primary target for a therapeutic intervention in AD, with numerous phase III trials underway or completed and designed to reduce Ab levels using a range of strategies. Thus far, the results are either negative or inconclusive, leading many to call into question the underlying validity of the amyloid hypothesis.